Herpes simplex virus type 1 (HSV-1) cause two modes of infections in humans, productive and latent. In the course of a productive infection of epithelial cells, the virus is taken up into axonal terminals and spreads within the axon by retrograde axonal flow to the nucleus in the body of sensory ganglia, such as the trigeminal or sacral ganglion. Some sensory neurons undergo productive infection and are destroyed, as are the cells at the site of entry. In other sensory neurons the virus establishes a latent infection and can reside in a nonproductive state for the rest of the life of the host. From time to time, however, the virus can be reactivated, triggered by local stimuli (injury) or by systemic conditions (UV-light, fever, stress), causing productive infection in epithelial cells near the site of entry. In latently infected neurons, viral DNA is found as an episome condensed into a chromatin-like structure. According to the two kinds of infection, HSV-1 can express two distinct transcriptional patterns, one specifying productive infection and the other specifying latency. Productive infection which is typical at peripheral sites in the mammalian host or during encephalitis involves a regulated cascade of gene expression. Immediate-early (IE) genes are expressed first and are required for the efficient expression of the subsequent early and late genes. The productive cascade generates the components required for viral replication and the production of infectious progeny. In contrast, during latency the virus is in a relative quiescent state, the expression of the more than 75 genes of HSV-1 is repressed and transcription is restricted to a single region of the viral genome coding for the latency-associated transcripts (LATs). The LATs which are not abundant during productive infection accumulate in the nuclei of latently infected neurons. Neither has a protein product been associated with the LATs nor a function defined for the LATs. In animal models virus mutants which do not transcribe the lat-gene are nonetheless able to establish and maintain latency, although some of these mutants are defective in reactivation from latency. The absence of infectious virus and detectable viral antigens correlates with the absence of detectable transcripts from productive-cycle genes in human trigeminal or sacral ganglia infected with HSV-1. However, infectious virus and viral products can be readily detected during the acute phase of establishment of latency and during reactivation. The mechanisms by which the productive pattern of gene expression in acutely infected neurons switches to the latent pattern, and how productive transcription is reactivated remain central unanswered questions in HSV biology, with significance for the control of HSV disease. It has been proposed that the absence of IE expression permits the establishment and maintenance of latency, and the activation of IE genes would lead to reactivation. Although HSV-1 is responsible for a viral encephalitis with the highest fatality rate, much less is known about the molecular biology of latent HSV-1 infection in the CNS than in the PNS. In animals it could be shown that, following viral replication within the trigeminal ganglion, the virus can travel and replicate in the CNS. However, the source of viruses causing herpes encephalitis is not always known. It has been suggested that encephalitis is due to HSV-1 reactivation from trigeminal ganglia, since the infectious process tends to involve the temporal and frontal lobes, brain regions with blood vessels and meninges which derive their sensory innervation from TG. Moreover, since not all cases of viral encephalitis are caused by the same viral strain that is responsible for cold sores in the individual, it is assumed that in nearly half of the patients HSV-1 encephalitis occurs during primary infection on the background of a preexisting latent infection.
The DNA methylation pattern is a hallmark of mammalian genomes. The DNA methyltransferase (MeTase) is responsible for methylation at the 5-position of cytosine residues located in the dinucleotide context Cytosine-Guanidine (CpG). Only twenty percent of the CpG sites are nonmethylated, and these sites are distributed in a nonrandom manner to generate a pattern of methylation that is site-, gene-, cell-, and tissue-specific. The specific patterns of DNA methylation that are seen in the adult organism are established in a programmed manner during development which results from the combination of maintenance/ de novo methylation, demodification of the methylated CpGs. As a result of these events, the final adult modification pattern is characterized by full methylation of those genes that are inactive and undermethylation of those genes that are active. The DNA MeTase activity is tightly regulated with the growth state of cells and previous models have suggested that the level of DNA MeTase activity can play an important role in determining the pattern of methylation. Changes in the pattern of DNA methylation have been correlated with a number of physiological processes in mammals. These include, besides differentiation, somatic X-chromosomal inactivation in females, genomic imprinting, formation of chromatin structure, the timing of DNA replication. Whether DNA methylation is the cause or the consequence of these processes still remains to be resolved. Nevertheless, it is clear that DNA methylation is an essential process, at least in mammalian development, since mouse embryos deficient for the known DNA methyltransferase do not survive past midgestation. Furthermore, the methylation pattern seems to be not only dynamic under physiological but also under pathological conditions, such as cancer pathogenesis. Aberrant methylation also has been demonstrated in the fragile X-syndrome for the fmr-1 gene.
The primary role of DNA methylation seems to be the long-term repression of gene expression. The influence of DNA modification on gene transcription is probably mediated by an effect of methyl moieties on DNA-protein interaction. For many genes, the major target of this modification appears to be the upstream regulatory region. Two ways of transcriptional silencing are known. Firstly direct, by preventing the binding of transcription factors to the promotor. Secondly indirect, by inhibiting the interaction of transcription factors with the promotor through certain methylated DNA-binding proteins. DNA methylation can be viewed as a way of dividing the genome into a small, unmethylated compartment that is accessible to diffusible regulatory factors, and a large methylated compartment that is propagated in the condensed state. This compartmentalization is thought to reduce the time required for regulatory proteins to locate their target sites in DNA and to provide an "immune" function for the rest of the genome by restricting the targets of invasive sequences such as viral DNA. In several in vivo systems, including adenoviruses and retroviruses, latent or inactive viral genomes are hypermethylated, wheras replicating virus are consistently hypomethylated. For HSV-1, methylation of the viral genome in a latently infected cell culture could be demonstrated. In contrast, in vivo the HSV-1 DNA present in the CNS of latently infected mice appeared not to be extensively methylated, using an insensitive method.
The aim of this study is to examine with a sensitive method if and subsequently how DNA methylation might contribute to the establishment and maintenance of HSV-1 latency and to assess its relevance to human herpes simplex disease. At first, a sensitive method to screen a large part of the HSV-1 genome will be established. The procedure is based on a method described by Frommer et al. (Proc. Natl. Acad. Sci. USA 1992 89, 1827-31). Following Frommer the genomic DNA is modified using bisulfite under conditions where cytosine is converted to uracil and 5-methylcytosine remains unchanged. The target sequence is then exponentially amplified using polymerase chain reactions (PCR) and strand-specific primers. Differing from the Frommer method, the amplified DNA fragment will be 10-20 times larger (5-10 kb) and the methylated loci will be detected by ddC-polymerase extension, exonucleolyse and agarose gel ectrophoresis. For methylation, screening DNA is extracted from homogenized autopsied human trigeminal ganglia which carry the HSV-1 genome in more than 80% of individuals. The DNA for the control is purified from acute infection, either from the CSF of patients with herpes encephalitis or from tissue of mucocutaneus lesions. For methylation analyses we have choosen the part of the 152 kb long HSV-1 genome encoding 3 of 5 IE genes and the LATs as well as the neurovirulence gene.